Fri, 01/14/2022 - 19:36
I’ll start with several statements:
- If you’re using a DSLR or one-shot color CMOS (OSC) camera, photometric measurements are reported as TG, TB, and TR. Observations are calibrated with a V-band, B-band, or R-Band comp star, respectively.
- The use of transformation equations is typically done for correcting for equipment-specific variables.
- It is possible to use transformation equations to transform between photometric systems such as Johnson & Cousins and Sloan. These transformation equations deal with both the change of system and equipment-specific issues.
Assuming these statements are correct, obviously, transformation equations can be developed between Tri-color derived results and the J&C standard. If these transformations can be done with sufficient accuracy, could the results be reported to the AAVSO as V, B, R with TRANS=YES, including, perhaps, an additional comment in the NOTES section? What might be a sufficient accuracy? Are there any benefits to reporting in this way, again assuming accuracy is sufficient?
Or, would the AAVSO prefer DSLR/OSC-derived results always reported as tri-color to avoid any confusion in the interpretation of said results?
Gary,
Yes, DSLR images can and should be transformed to J/C and reported as J/C.
For now software like VPhot do not make the transformation very easy as you have to play naming games when describing your inputs. But it can be done and the CHOICE courses cover the workarounds.
We are working on a naming pattern for these transform coefficients so that it will be clear what kind of transform you are applying. Something like tgTv_bv would indicate a transform of TG instrumental data to J/C V.
Adding information in the NOTES fields is always welcome!
George
So taking input from both replies, is the following correct:
"Yes, DSLR images can and should be transformed to J/C and reported as J/C."
DSLR means, DSLR, OSC, or CMOS mono with RGB filters
You should submit transformed results if you believe the transformation to be accurate (see accuracy statement below).
"If you are submitting standardized magnitudes (i.e., non-transformed magnitudes) leave the "transformed" check box under the Magnitude field unchecked and select the appropriate "Tri-Color..." option from the Filter drop-down list. For example, select "Tri-Color Green" if the standardized magnitude was derived from green channel instrumental magnitudes."
I've developed transformation equations by measuring multiple standard fields comparison stars with an OSC CMOS camera. I can show RMS accuracy between 0.04 to 0.06 mags. Are these results sufficiently accurate to submit as transformed J&C, or should I submit as untransformed TG, TR, or TB?
"For transformed observations (with or without extinction correction} be sure to check the "transformed" check box and select either Johnson B, Johnson V or Cousins R from the Filter drop-down list, depending on which DSLR color channel was used."
Gary,
Yes, your summaries are correct.
In my opinion, if your standard deviations from repeated transformed determinations are 0.04 to 0.06 mags, then by all means submit them. I do not think you are likely to achieve that accuracy with non-transformed measurements unless the B-V of the variable and comparison stars are close to each other. When I was taking transformed measurements of 6th magnitude T Cen last year for a colleague to compare with spectroscopic results, SDs of 40 measurements (10 from each of 4 comp stars) were about 0.02 mag.
Roy
Hi Roy,
Yes, I should have added two limiting conditions:
I've drafted a paper on this that I've been reviewing with my AAVSO mentor. I should get this finished and submitted.
Gary
The last sentence of Gary's post reflects the potential confusion that can occur when DSLR photometry is reported. The relevant parts of the appropriate AAVSO publications are quoted, paraphrased and commented on below.
DSLR Observing Manual
Section 6.7 Submitting your results
"If you are submitting standardized magnitudes (i.e., non-transformed magnitudes) leave the "transformed" check box under the Magnitude field unchecked and select the appropriate "Tri-Color..." option from the Filter drop down list. For example select "Tri-Color Green" if the standardized magnitude was derived from green channel instrumental magnitudes."
"For transformed observations (with or without extinction correction} be sure to check the "transformed" check box and select either Johnson B, Johnson V or Cousins R from the Filter drop down list, depending on which DSLR color channel was used."
These instructions are very clear, however they conflict with the terminology in the AAVSO Extended File Format, as described below.
When this was discussed a while ago on the AAVSO Forum, after I pointed out this conflict, the answer from at least one experienced AAVSO observer was that the "Filter" field in the AAVSO Extended File Format refers not to the physical filter through which the observation was made, but to the passband of the submitted result.
AAVSO Extended File Format
There are a number of fields. I'll comment on those relevant to this discussion
OBSTYPE: If you as using a CCD camera or dedicated astronomical CMOS camera the OBSTYPE should be "CCD". If you are using a DSLR camera the OBSTYPE should be DSLR.
FILTER: There are of course many types, among them Johnson B, Johnson V, and Johnson R. For DSLR cameras, they are TG, TB and TR for Tri-color green, Tri-color blue and Tri-color R respectively. (Note: there is no mention of OSC - one shot colour - dedicated CMOS astronomical cameras here, but I assume the submission policy for data from these cameras should be the same as for DSLR cameras.)
TRANS: If the magnitude determination is transformed to a standard system (such as Johnson/Cousins) the entry in this field is YES. If the magnitudes are standardized magnitudes (i.e., not transformed to a standard system) the entry should be NO.
Conclusion
If you are submitting transformed DSLR photometry results and read only the instruction for the fields in the AAVSO Extended File Format, you will probably submit them incorrectly. It is necessary to read the relevant section of the AAVSO DSLR Observing Manual as well.
Roy
So, the Extended File format should contain the following information for the specified equipment configurations:
CCD mono with J&C filters OBSTYP = CCD, FILTER = B, V, R, etc. TRANS = NO, even if you transform for equipment differences according to the CCD Photometry Guide Section 6
CMOS mono with J&C filters OBSTYP = CCD, FILTER = B, V, R, etc. TRANS = NO, even if you transform for equipment differences according to the CCD Photometry Guide Section 6
DSLR OBSTYP = DSLR, FILTER = TB, TG, or TR. TRANS = NO
OR
DSLR OBSTYP = DSLR, FILTER = B, V, or R TRANS = YES; using the transformation equations for filter/equipment specified in the DSLR Handbook 1.4 Section 6.5
OSC CMOS OBSTYP = DSLR, FILTER = TB, TG, or TR. TRANS = NO
OR
OSC CMOS OBSTYP OBSTYP = DSLR, FILTER = B, V, or R TRANS = YES; sing the transformation equations for filter/equipment specified in the DSLR Handbook 1.4 Section 6.5
For DSLR or OSC, am I correct in staying the decision to transform is based on whether you believe the target star to contain emission or absorption bands. If it does, chose TB, TG, or TR. If it does not, ideally transform and submit as B, V, or R, other submit as TB, TG, or TR.
Gary:
Much of what you said is correct BUT why did you say that:
1. <<CCD mono with J&C filters OBSTYP = CCD, FILTER = B, V, R, etc. TRANS = NO, even if you transform for equipment differences according to the CCD Photometry Guide Section 6>>??
IF you use your J/C transformation coefficients to generate transformed standard BVRI magnitudes you would report TRANS=YES. Why do you think otherwise? There may be a small difference between non-transformed BVRI mags and transformed BVRI mags due to slight differences between your filters and the original filters used by Johnson / Cousins.
2. <<For DSLR or OSC, am I correct in staying the decision to transform is based on whether you believe the target star to contain emission or absorption bands. If it does, chose TB, TG, or TR. If it does not, ideally transform and submit as B, V, or R, other submit as TB, TG, or TR.>>
Generally yes. However, the TG to V transformation is quite good for most types of stars. TB and TR transformations are not quite as good and may be impacted more by emission /absorption lines that are more prominent in highly colored stars. You would be safe in just reporting as TB or TR but it is a tough call and J/C B or R may work okay?
You may test this by imaging a standard field and determining what differences you get in TR (non-transformed) vs R (transformed) mag for some highly colored blue or red comp stars or even some very red variables (tougher of course to compare to other's results) . Experimentation is better than just assumptions? Comments?
Ken
Relating to point 1, I made my comment because the description of the Extended File Format says:
TRANS: If the magnitude determination is transformed to a standard system (such as Johnson/Cousins) the entry in this field is YES. If the magnitudes are standardized magnitudes (i.e., not transformed to a standard system) the entry should be NO.
If you're using CCD/J&C, you are already in the standard system, you are not 'transforming to' it. However, if YES is the correct answer if you transform for equipment differences according to the CCD Photometry Guide Section 6, then I can see why that would be. I don't use CCD, so I was interpreting the instructions given for the Extended File Format. I did not read the CCD manual closely enough to ascertain if this point was covered there.
As for point 2 (since I use an OSC CMOS camera), I'm trying to clarify how I should submit, and what information to pass to my mentorees using similar equipment. While I have done quite a few measurements on standard star fields, I need to reprocess that data to determine differences between non-transform and transformed values, my analysis has concentrated on comparing only transformed to catalog values.
As you know, one of my interests is bringing new recruits into photometry. As some of these come from the world of astrophotography and EAA, they typically have RGB rather than BVR based systems. In terms of keeping things as simple as possible at the beginning, I would advise these new folks to initially submit as tri-color, and once they become familiar with transformation equations then consider submitting as BVR when appropriate. Would you agree this is the right approach?
Gary:
These two terms (standardized vs standard system) can be a bit ambiguous. The following is how I think about it:
I would first propose that virtually all magnitudes generated by aperture differential photometry and reported to the AAVSO are "standardized" since they have been calculated by comparison to known standard Johnson Cousins BVRI magnitudes documented in a comp star database.
Tri-color magnitudes from your OSC camera are calculated by comparison to the known comp star J/C magnitudes (TR from J/C R, TG from J/C V; TB from J/C B). Thus they are "Standardized" but I suspect you would agree that they are not on the "Standard System". The RGB filters/bandpasses are different enough from the standard J/C filters/bandpasses that you would not expect them to agree exactly, e.g., within say 0.01 mags. Due to these differences, AAVSO decided to report them as TG not V, etc. If you want to report them as V, you must transform them to the "Standard System". This transformation corrects the tri-color standardized magnitudes to the Standard J/C System.
The same thing can be said for a CCD camera with J/C filters/bandpasses. The difference in this case is that most Johnson/Cousins filters manufactured and purchased by amateurs today are quite close BUT not exactly the same as those originally designed/fabricated/used by Johnson and Cousins. They are closer than the tricolor filters but even they may not be exactly the same (i.e., identical) as the original J/C filters. So again, the magnitudes are "Standardized" but not on the "Standard System". Transformation is again needed to correct the BVRI standardized magnitudes to the "Standard J/C System".
As you can imagine, the transformation coefficients to correct your J/C filter magnitudes to standard J/C filter magnitudes are much closer to the ideal 1 and 0 values than for Tricolor filter magnitudes to standard J/C filter magnitudes.
So, for better or worse, that's how I look at the wording ambiguity! Do you buy it?
Ken
Thanks for clarifying the correct approach to submitting. To summarize:
CCD mono with J&C filters OBSTYP = CCD, FILTER = B, V, R, etc. TRANS = NO, if NOT transformed for equipment differences according to the CCD Photometry Guide Section 6
OR
CCD mono with J&C filters OBSTYP = CCD, FILTER = B, V, R, etc. TRANS = YES, if transformed for equipment differences according to the CCD Photometry Guide Section 6
CMOS mono with J&C filters OBSTYP = CCD, FILTER = B, V, R, etc. TRANS = NO, if NOT transformed for equipment differences according to the CCD Photometry Guide Section 6
OR
CMOS mono with J&C filters OBSTYP = CCD, FILTER = B, V, R, etc. TRANS = YES, if transformed for equipment differences according to the CCD Photometry Guide Section 6
DSLR OBSTYP = DSLR, FILTER = TB, TG, or TR. TRANS = NO; if NOT transformed
OR
DSLR OBSTYP = DSLR, FILTER = B, V, or R TRANS = YES; using the transformation equations for filter/equipment specified in the DSLR Handbook 1.4 Section 6.5
OSC CMOS OBSTYP = DSLR, FILTER = TB, TG, or TR. TRANS = NO; if NOT transformed
OR
OSC CMOS OBSTYP OBSTYP = DSLR, FILTER = B, V, or R TRANS = YES; using the transformation equations for filter/equipment specified in the DSLR Handbook 1.4 Section 6.5
Hopefully, I got it correct. As you said it's somewhat ambiguous, so if I'm confused am I sure others are.
In an earlier post in this thread, I wrote (in part):
"Conclusion
If you are submitting transformed DSLR photometry results and read only the instruction for the fields in the AAVSO Extended File Format, you will probably submit them incorrectly. It is necessary to read the relevant section of the AAVSO DSLR Observing Manual as well."
This has been the situation for a few years.
In my opinion, the instructions for the Extended File Format should have been updated to incorporate the reporting of transformed magnitudes from DSLR cameras, OSC astro cameras and mono CCD or CMOS astro cameras used with astrophotography filters.
Roy
So Roy is my summary correct?
Best regards,
Gary
Gary, the only points that should I think be different are for the CMOS dedicated astro cameras. I think the AAVSO convention for these is that they should be designated as CCD not DSLR.
The experts are on this list. They will correct me if the above is wrong.
Personally, I think it would be best for HQ to create a new Extended File Format category CMOS ASTRO for dedicated CMOS astrophotography cameras instead of using the counterintuitive designation CCD for a sensor that is CMOS. But that's just my opinion.
Roy
I think color CMOS, or for that matter mono CMOS (or even CCD) using RGB rather than BVR filters should be designated DSLR. The DSLR is not really describing the camera type, rather that the results are coming from an RGB-based system.
If you specify a color CMOS as CCD, and submit non-transformed results as TR, TG, or TB then this indicates this comes from an RBG-based system. But if you transform and submit as CCD, BVR, TRANS=YES, how can you tell this comes from an RGB-based system rather than a BVR-based system unless you make a comment in the comment section?
Gary, I don't make the AAVSO rules. I just try to understand what they are.
You have touched on an issue which has been discussed previously on the Forum - the issue of filters versus passbands.
TR, RG, TB refer to non-transformed data from an RGB-based system.
When data from an RGB camera is transformed to the Johnson system (Cousins for R), the AAVSO rule is that the result is reported as the filter that describes the passband. Thus, if you transform data from the green channel of an RGB camera (whether it is a DSLR or a colour CMOS camera) to the Johnson V passband, the data is reported as 'transformed' and as 'V'. This is the rule. It is enunciated in the AAVSO DSLR Photometry Observing Manual. Data reported thus describes a passband, not a filter used.
Roy
Gary:
Your summary is correct. Currently, CMOS and OSC cameras do not have their own designations in webobs. Therefore, OSC is reported as DSLR, and CMOS is reported as CCD.
The assumption is that the One Shot Color (OSC) camera and DSLR camera both have Bayer arrays. The main difference between OSC and DSLR is that the OSC is packaged differently and has thermal cooling. However, they are both just reported as DSLR. Note that in most/all cases, today's OSC and DSLR cameras happen to use CMOS sensors. The key is that they are Bayer array color cameras.
The assumption is that the CMOS and CCD are monochrome cameras. They are both just reported as CCD. If they used J/C filters, the mags would be reported as BVR as you indicted above.
(Note, obviously IF the two assumptions above were not true for some unexpected reason, the rule would no longer be true. I have not seen any of these unexpected circumstances?)
Note that it is possible for one of the CMOS/CCD monochrome cameras to have used RGB astrophotography filters as opposed to J/C filters. In this case, the non-transformed magnitudes would be reported as TB, TG, TR. The transformed mags would be reported as BVR.
Note that it is also possible for one of the CMOS/CCD monochrome cameras to have used Sloan (SG, SR, SI) filters. In this case, the mags would be reported as SG, SR, SI, whether reported as transformed or non-transformed. However, note that there are historically no Sloan magnitudes automatically available for comparison stars. This is about to change! Sloan mags will be available for comp stars in VSP soon. The rule will be only to transform Sloan mags to Sloan mags, not J/C mags.
Ken
Thanks, Ken. I'm glad I've got this correct now :).